By Ana Arcanjo
January 2025 marked a defining moment in the ongoing debate about AI-generated art and copyright protection. The U.S. Copyright Office, after reviewing more than 10,000 public comments, released its long-awaited report confirming what many creators have suspected: AI art can only be copyrighted if there’s a meaningful human contribution involved. In other words, if the machine is doing all the painting, don’t expect legal protection. But if you’re still driving the creative wheel, even with AI as your co-pilot, then your rights may stand a chance.
This ruling not only responds to a growing cultural concern but directly intersects with the questions I explored in my master’s thesis: what happens to authorship when creation becomes co-authorship with a machine? This blog post blends the official stance of the U.S. Copyright Office with my research into human-machine creativity to offer a critical, reflective, and slightly provocative take on where we’re headed.
According to the U.S. Copyright Office’s 2025 report, copyright depends on human expression. Prompts alone, or letting a generative system like Midjourney or DALL·E do the work from start to finish, do not meet the threshold for legal authorship. What’s protected is the creative selection, coordination, or modification that reflects human intentionality. The key? Proving your involvement.
No new laws were introduced; the existing ones are being interpreted through a new lens. And that lens is laser-focused on one idea: authorship is human.
A key illustration from my thesis was the case study of Student A, who explored her creative influences and aesthetics through a hybrid design process using AI tools. Her work revealed a thoughtful, layered construction of visual identity that integrated references ranging from Taylor Swift’s Folklore to the architecture of Norman Foster and Gaudí.
image from student A trial
In her process, AI was not the author—it was a medium through which she experimented with composition, palette, and mood. As seen in the images, she curated AI-generated imagery to reflect personal aesthetics, incorporating human judgment at every stage. The expressive coherence across references and outputs confirmed that the artist’s vision shaped the final result, not the algorithm alone.
This supports the Copyright Office’s position: AI may assist, but authorship demands demonstrable human input. Student A’s work proves that even in AI-supported creation, the artist’s role remains central and legally recognizable.
In my thesis, I defended the view that the artistic process is where authorship is formed, not just the final image. The click of a prompt doesn’t constitute creativity, but curating outputs, iterating based on intuition, and reworking fragments does. AI, I argued, doesn’t replace the artist. It expands their reach, challenges their habits, and reconfigures their workflows.
My data showed that creators who integrated AI into their practice—not as a shortcut but as a partner—still made decisions that were deeply expressive and unique. In one interview, a visual artist described their use of AI as a “collaborative sketchpad”—never the final brushstroke, but always present in the early sparks.
This human-centered, process-oriented interpretation resonates with the Copyright Office’s notion of AI as a tool, not a stand-in. The distinction isn’t what the AI can do, but how the human uses it.
The report underlines a crucial nuance: using AI to refer to something is okay. But using AI to generate the entire piece? Not enough. We must be able to show that we controlled or shaped the expressive elements. This is less about how smart the machine is and more about how present the artist was.
And here’s where things get murky. Because “presence” in digital creativity isn’t always obvious. When I analyzed case studies for my dissertation, I found that even slight changes in prompt phrasing could lead to drastically different outputs. Artists weren’t just writing—they were sculpting language into vision. Can that really be dismissed as passive input?
What this ruling demands is not the end of AI in art, but a clearer articulation of human intent. It forces artists to ask: where am I in this work? That’s not a constraint—it’s an invitation. My research showed that the artists who embraced this question became more critical, more intentional, and paradoxically, more human in their work.
Of course, legal gray areas remain. Will we need metadata trails to prove authorship? Will AI companies own part of our visions? Will generative tools develop features to log creative decisions? These are unresolved questions. But one thing is certain: the narrative of authorship is shifting, and creators must shape it, not just react to it.
We’re in a moment of cultural negotiation. The image might be born in code, but its soul still comes from human curiosity, judgment, and experimentation. If the U.S. Copyright Office tells us anything, it’s this: the future of art isn’t about AI replacing artists—it’s about making sure artists still matter when AI is part of the process.
Interior Architect, MSc, in Design, and Interdisciplinary Researcher. Enthusiast of innovative teaching methodologies.
Learn more about Ana Arcanjo: Instagram * Linked In
